A Gas Pressure Scale Based on Primary Standard Piston Gauges

Accurate determinations of pressure are required for diverse applications such as the altitude of aircraft, barometric pressure for weather prediction, and process control in manufacturing.  The pressure standard for NIST in the atmospheric pressure range is the mercury manometer, which is the most accurate pressure standard in the world but has an upper limit in pressure of 350 kPa.  The purpose of this work is to revise the gas pressure scale based on piston gauges which overlaps the mercury manometer and extends up to a pressure of 17 MPa.  This gas pressure scale utilizes a suite of piston gauges that are traceable to dimensionally-characterized piston gauges, and produces significant reductions in pressure uncertainties that NIST can pass on to its customers.

A piston gauge is a round piston fitted in a matching cylinder; the piston is loaded with weights of known mass and density.  The piston is marginally smaller than the cylinder, and fluid fills the gap between the two components.  In 2006, NIST established two, 36 mm diameter piston-cylinder artifacts (referred to as PG38 and PG39) as primary standards through careful dimensional characterization of their diameters, modeling of the forces acting on the artifacts, and experimental comparison of the artifacts to each other and to the NIST mercury manometer.  Establishing a pressure scale based on the primary standards required careful intercomparisons between those primary standards and ten secondary standard piston gauges.  The secondary standard piston gauges have successively smaller diameters to allow extending the range from 1 MPa (the upper limit of PG38 and PG39) up to 17 MPa.

A pressure scale traceable to primary standard piston gauges provides two significant advantages over a scale traceable to the mercury manometer.  The first is that the upper limit of 1 MPa for PG38 and PG39 allows determining pressure-induced distortion in the secondary standard piston gauges, which is not possible using the mercury manometer.  Distortion is a major contributor to uncertainty at higher pressures.  The second advantage is that calibrations against piston gauges require much less time than that against manometers, making it feasible to perform more comparisons of the secondary standards to PG38 and PG39.

This work established a rigorous basis for the piston gauge effective area and uncertainty.  As examples of reductions in expanded (k=2)  uncertainty that were achieved, at 350 kPa the uncertainty dropped from 19 ppm to 10 ppm; at 1.4 MPa the uncertainty dropped from 33 ppm to 13 ppm; and at 7 MPa the uncertainty dropped from 33 ppm to 19 ppm.   New Calibration Measurement Capabilities were approved by SIM and BIPM in 2008 and are available to NIST customers.

Presently, PG38 and PG39 can only be operated in gauge mode. In the future, we will extend the comparison capability of the primary standards to absolute mode.  We will also further propagate the pressure scale to 100 MPa for gas and 280 MPa for oil.